| 利用神经网络进行复杂自由曲线／曲面重构 |
| |
| 引用本文： | 史桂蓉,邢渊,张永清. 利用神经网络进行复杂自由曲线／曲面重构[J]. 上海交通大学学报, 2000, 34(10): 1372-1375 |
| |
| 作者姓名： | 史桂蓉 邢渊 张永清 |
| |
| 作者单位： | 上海交通大学国家模具CAD工程研究中心，上海 200030 |
| |
| 摘 要： | 提出了利用Ｂｅｚｉｅｒ基函数神经网络重构自由曲面的方法。该方法能直接从神经网络的权值矩阵得到曲线的控制顶点／曲面的控制网络，通过神经网络的权值约束实现曲线段／曲面片之间的Ｇ＾１０、Ｇ＾１光滑拼接，给出了神经网络的构造学习过程，并用实例加以验证。
|
| 关 键 词： | 神经网络 曲线 曲面重构 CAD |
| 文章编号： | 1006-2467(2000)10-1372-04 |
| 修稿时间： | 1999-12-17 |
Reconstructing Free-Form Curve/Surface Using Bezier Basis Function Neural Network |
| |
| SHI Gui-rong,XING Yuan,ZHANG Yong-qing. Reconstructing Free-Form Curve/Surface Using Bezier Basis Function Neural Network[J]. Journal of Shanghai Jiaotong University, 2000, 34(10): 1372-1375 |
| |
| Authors: | SHI Gui-rong XING Yuan ZHANG Yong-qing |
| |
| Abstract: | The free- form surface reconstruction is one main objectof CAD/CAM,and itis a key step in re- verse engineering.This paper presented a new approach to free- form surface reconstruction using Bezier basis function neural networks( BBFN) .Advantages of thismethod are asfollows:getting directly control vertes of free- form curve/surface from weights of BBFN;providing G0 ,G1 connecting among multi- patches by constraining weight- ajusting between multi- BBFNs.The paper described in detail the structure and learning program of BBFN,and presented several examples. |
| |
| Keywords: | neural network curve curved surface surface reconstruct reverse engineering |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
